Risk, uncertainty and knowledge

Abstract

While the development of modern medicine is associated with both increases in scientific knowledge and improved outcomes in health care it is also associated with increased uncertainty as expert and lay knowledge bases have diverged and separated. The development of a principal-agent relationship in the late nineteenth and early twentieth century in which medical practitioners used their specialist knowledge to make decisions for and on behalf of their patients provided one way of managing this uncertainty. However the development of a less deferential and more consumerist culture associated with medical scandals in which trust has been betrayed have led to increased regulation of medical practice, especially the development of national standards based on encoded knowledge. Even if governments can overcome the practical problems of using such systems to structure decision-making, because these systems fail to address the personal and emotional components of trust they are likely to create a ‘trust deficit,’ a system that may work better, but is trusted less.

Keywords:

• Knowledge
• risk
• uncertainty
• trust
• rationality
• affect
• emotion

Introduction

Modern science has created a sophisticated body of knowledge of the causes and most effective ways of treating diseases. For example, research programmes such as the human genome programme have deciphered the human genetic code, facilitating the identification of genetic errors associated with breast and cervical cancer (Human Genome Project Information n.d.), and the development of disease specific research networks (Department of Health 2006) means that new treatment regimes can be rapidly tested, enhancing treatment outcomes. Such developments have had clear measurable benefits. In 1970 approximately 50% of women diagnosed with breast cancer survived 5 years; currently 80% survive (Lyall 2006: 1).

However such objective population-wide improvement in health and well-being is not reflected in individual subjective assessments of health and security:

Material levels of security in the western world are higher than ever before … However, the sources of uncertainty and the mechanisms available to most people to deal with them have changed, leading to the paradox of timid prosperity—growing uncertainty amid rising affluence (Taylor-Gooby 2000: 3).

Uncertainty and anxiety are especially associated with illness and disease, whether this is the threat of pandemic disease such as SARS or the more personal experience of chronic illness (Locker 2003). Science is seen as a source of knowledge which paradoxically both reduces and contributes to uncertainty. While science and the professions that applied science to practical issues, such as medicine, commanded considerable public support in the early twentieth century, events of the late twentieth century have tended to undermine this authority. Moran describes the ways in which in the UK the BSE Disaster and subsequent Inquiry exposed the limitations of scientific knowledge and ‘was part of a larger crisis of confidence in the food safety regime that produced periodic food scares’ (Moran 2003: 149). The Inquiry into BSE examined the ways in which the government handled a phenomenon which was ‘a known hazard to cattle and an unknown hazard to humans’ (BSE Inquiry 2000: 1). The Inquiry noted that the Ministry of Agriculture, Food and Fisheries relied heavily on ‘the advice of independent scientific experts.’ While the Inquiry did not feel that the government sought to mislead the public, its misplaced attempts to reassure the public backfired, undermining public confidence. Though such mistakes may emphasize the limitations of current knowledge and undermine confidence, there are other more fundamental problems in applying and using knowledge in contemporary societies.

In pre-modern society knowledge systems, especially those used to explain and manage threats to collective and individuals' well-being such as illness, tend to incorporate religious and supernatural systems. Such understandings provide both an explanation of past events, i.e. why a particular individual died, and also a means of predicting the future and avoiding hazards. For example Evans-Pritchard (1976), in his classic study of a traditional society in Southern Sudan, demonstrated the ways in which the Azande could explain all misfortune, often combining empirical evidence, e.g. termites eating through the wood support of granary or the skill of a particular potter, with religious ideas such as witchcraft or sorcery to explain why a particular individuals was sitting under a granary when it collapsed or a particular batch of pots misfired. They used the same knowledge to predict the future.

Such systems explain everything (and therefore to modern minds nothing) and have no space for uncertainty. In contrast the modern scientific theories and knowledge that have come to replace such belief systems are essentially limited. They are an attempt to capture, explain and approximate reality. As Ilkka notes ‘good scientific theories are typically false but nevertheless “close to the truth”’ (Ilkka 2002: vi). The limitations of a theory become evident when it is replaced by new theory. For example the inadequacies of the ‘humour’ theory of the body, which underpinned medical practice until the seventeenth century, were exposed by the conceptualization of the body as a system of interconnected organs. However at any particular time it is difficult to be sure how ‘close to the truth’ accepted theories are and therefore scientific knowledge is always provisional and uncertain. For Beck this uncertainty underpins contemporary concerns with risk:

Risk society begins where tradition ends, when, in all spheres of life, we can no longer take traditional certainties for granted. The less we rely on traditional securities, the more risks we have to negotiate. The more risks, the more decisions and choices we have to make (Beck 1998: 10).

The growth of uncertainty

While science has considerably enhanced knowledge of the natural world, it does not necessarily provide the type of information by which individuals can manage their lives. For instance, epidemiology provides information on the factors associated with the development of disease based on a mapping of the incidence of diseases in defined circumstances. Publication of the first findings in 1950 that smoking was linked to lung cancer (Doll and Hill 1950, Wynder and Graham 1950), and the consensus is that about half of regular cigarette smokers will be killed by the habit (Peto 1994). However, an individual smoker can't know for sure which half they are in.

The growth and increasing fragmentation of scientific knowledge into specialist disciplines creates an additional area of uncertainty. As experts become increasingly knowledgeable and specialized, so the gap between their knowledge and that of non-experts increases. Thus while consultations between Azande experts and their clients are based on the assumption that they share the same knowledge and beliefs, consultations in late-modern societies lack such universalized validity claims (Habermas 1987). A crucial part of the interaction is ‘communicating’ or translating between different bodies of knowledge. Morgan suggests that experts such as doctors can either seek to impose their knowledge by focussing ‘on objective descriptions of symptoms … within a reductionist biomedical model' or seek to translate and adjust to “patients” own illness framework’ (Morgan 2003: 55). Nettleton also notes the ways in which scientific knowledge is challenged by other forms of knowledge such as that derived from personal experience and the ways in which this undermines experts own confidence in such knowledge (Nettleton 1995: 33). Knowledge development thus undoubtedly creates social benefits in modern society, for example a healthier population living longer, but this knowledge does not reduce uncertainty, indeed for a variety of reasons, it actually increases it.

The agent relationship as a response to uncertainty

Individuals or groups may seek alternative sources of knowledge to overcome the limitations of scientific knowledge (for example, through alternative medicine), however there is an officially sanctioned alternative, the principal-agent relationship in which the agent ‘acts on behalf of another even though there are differences of interest and inequalities of information between them’ (Barbalet 2006: 7). This relationship reduces the ‘potentially high costs associated with the actual process of decision-making and those associated with making the wrong decision (i.e. anxiety costs)’ (McGuire et al. 1988: 186). However this approach can only work effectively when there is trust, i.e. the individual believes that the agent will act in good faith. Trust can be defined as the as ‘confident expectations about another's motives with respect to oneself in situations entailing risk’ (Boon and Holmes 1991: 194) and is ‘the means whereby an uncertain future is given the semblance of certainty’ (Barbalet 2006: 7).

The establishment of the NHS in the UK in 1948 was centred on the doctor – patient relationship. The aim was to give doctors the resources and autonomy to identify and meet all patient need for physical and mental healthcare. Thus the government entrusted the profession with the provision of ‘comprehensive medical care and treatment’ (Webster 1988: 112) and accorded the profession its ‘traditional freedom and dignity,’ expecting the population to trust their doctors. While the doctor – patient relationship and the agency structure which underpins it remains central to the delivery of health, the nature of this structure has changed markedly.

Informed consent is a relatively recent phenomenon in medicine. In the early nineteenth century there was an ethical debate within healthcare in which Percival, in a text on medical ethics first published in 1803 (Leake 1927), argued that it would be a ‘gross and unfeeling wrong to reveal the truth’ to a patient if such truth was harmful. His friend the Reverend Thomas Gisborne dissented arguing in 1794 that: ‘The physician … is invariably bound never to represent uncertainty or danger as less than he actually believes it to be.’ In the nineteenth century Percival's arguments prevailed and as the American College of Obstetricians and Gynecologists ethics committee noted until the 1970s medical practice was grounded in paternalism with doctors deciding how much information to give their patients using ‘the medical well-being of the patient’ (American College of Obstetricians and Gynecologists 2004: 10) to justify their decisions.

In the 1970s there was a ‘paradigm shift’ (American College of Obstetricians and Gynecologists 2004: 9) from a ‘protective’ approach to one in which the patient has the right to enough information to make an informed decision and give informed consent to treatment. The doctor must effectively communicate all medically relevant knowledge including the risks and uncertainties. The American College of Obstetricians and Gynecologists states that the patient should be:

given adequate information about her diagnosis, prognosis, and alternative treatment choices, including the option of no treatment (American College of Obstetricians and Gynecologists 2004: 10).

The development of informed consent involves a major shift in medical practice. The doctor can no longer make decisions on behalf of the patient, i.e. act as agent, but becomes an adviser or counsellor who provides knowledge about both the benefits and risks of treatment so that the patient can make an informed choice. Thus the doctor has a duty to effectively communicate information on risks.

This change in medical practice and medical roles was the product of a combination of factors including the increased involvement of the law in medical cases, a decline in deference, the rise of consumerism in post-industrial democracies and government responses to medical scandals. Dissatisfied patients have always had recourse to the law. For example in eighteenth century England a patient sued two doctors who had rebroken and reset his fractured leg without his consent (Slater v Baker and Stapleton, K. B. 1767, cited in Jester 1998: 1). In the twentieth century informed consent became the basis of legal judgements. In the USA, Justice Cardozo presided over a case in which a woman had consented to an abdominal examination under anaesthetic but not an operation and during the operation the doctor had found and removed a tumour. In a landmark decision the Judge found in the patient's favour ruling ‘that an individual had the right of bodily self-determination’ (Schoendorff v Society of New York Hospital, NY 1914, cited in Jester 1998: 1).

Medical scandals and the rationalization of healthcare

Medical scandals in the UK, particularly at Royal Liverpool Children's Hospital and the Bristol Royal Infirmary, have also contributed to changing conceptions of doctor – patient relationship and have provided a pretext for government to increase the regulation of medical practice. In 1999 the government established a public inquiry into treatment of dead children's bodies at Alder Hey Hospital. The Inquiry found a ‘long-standing widespread practice of organ retention without consent’ (The Royal Liverpool Children's Inquiry 2001: 444). The Inquiry felt that these practices were unacceptable and reflected an outdated approach to medical ethics and practice that should be consigned ‘to the annals of history’ (The Royal Liverpool Children's Inquiry 2001: 445):

Their [the medical profession] approach has been paternalistic in the belief that parents or relatives would not wish to know about the retention of organs and the use to which they were put … In the current climate of frankness and openness it should no longer be possible for organs to be retained without the knowledge or consent of the parents (The Royal Liverpool Children's Inquiry 2001: 4).

The policy response focused mainly on the regulation of medical and other health research with the development of research ethics and governance framework. The Bristol Royal Infirmary Inquiry examined the services provided by the paediatric cardiac surgical team at Bristol Royal Infirmary between 1984 and 1995. The Inquiry found that the service was so poor that it exposed children under the age of one to an unacceptable level of risk or ‘beyond those ordinarily to be expected in the time and context’ (Bristol Royal Infirmary Inquiry 2001: 131) and estimated that from 1991 until 1995 between 30 and 35 children died who would not if the quality of services at Bristol had been comparable to that of other regional centres (Bristol Royal Infirmary Inquiry 2001: 4). As at Alder Hey there was a problem of informed consent, at Bristol parents were not aware of the risks to which their children were being exposed. For example Michael Parsons, whose daughter Mia died following cardiac surgery at Bristol Royal Infirmary, stated that:

Had we known of the real statistics we would never have accepted the referral … When I signed the consent form I believed I was doing the best thing for Mia. However, in retrospect I know that I did not. I maintain that my consent was obtained by giving me false information. This is in my view criminal (quoted in Boseley 1999).

The Inquiry found that a club culture had developed at the hospital in which a core group of managers disregarded information and withheld held it not only from parents but also from other professionals, managers and the public:

Openness: Bristol was awash with data. There was enough information from the late 1980s onwards to cause questions about mortality rates to be raised both in Bristol and elsewhere had the mindset to do so existed. Little, if any, of this information was available to the parents or to the public. Such information as was given to parents was often partial, confusing and unclear. For the future, there must be openness about clinical performance. Patients should be able to gain access to information about the relative performance of a hospital, or a particular service or consultant unit (The Bristol Royal Infirmary Inquiry 2001: Synopsis, para 18).

The Inquiry drew on and stimulated contemporary government thinking linking enhanced medical regulation and clinical governance, i.e. increased accountability for medical decision-making. These changes are designed to enhance public confidence in heath care through a process of quality assurance which linked openness to self and external regulation:

Clinical governance will be the process by which each part of the NHS quality-assures its clinical decisions … Professional self-regulation provides clinicians with the opportunity to help set standards. People need to be confident that the regulatory bodies will exercise rigorous self-regulation over the standard and conduct of health professionals and will act promptly and openly when things go wrong (NHS Executive 1999: 2 – 3, emphasis added).

From tacit to encoded knowledge

This new approach to regulation is based on a critique of the ways in which knowledge is used in the NHS. As Barbalet notes: ‘considerations of whether professionals might be trusted … typically concern belief about the adequacy of a knowledge base or its application (Barbalet 2006: 8). The Bristol Inquiry identified a ‘club culture’ (The Bristol Royal Infirmary Inquiry 2001: Synopsis, para 8) in the hospital that insulated internal practises from external scrutiny based on national standards. Thus the new system of regulation is designed to replace local knowledge based either on group custom and practice or personal intuition with national standards based on scientific evidence. Thus in the ‘New’ NHS, national systems for encoding knowledge in guidelines and national frameworks provide the basis for the regulation and scrutiny of the delivery of healthcare.

While the NHS and related research programmes are designed to provide the basic knowledge or evidence, a variety of Institutions, which are supported by the Department of Health, such as the Cochrane Collaboration and National Institute for Health and Clinical Excellence (NICE) summarize and codify this evidence. The Cochrane Collaboration (2006) sees itself as ‘a reliable source of evidence in healthcare’ while NICE emphasizes its role as:

an independent organization responsible for providing national guidance on promoting good health and preventing and treating ill health … [including] guidance on the appropriate treatment and care of people with specific diseases and conditions within the NHS (NICE, 2006).

This codified knowledge provides the framework of clinical decision-making. Doctors are expected to use it and managers are expected to check that it is being used. The main source of evidence for compliance with national standards is the outcome of clinical practice. The 1997 White Paper sees outcome measurement as the basis of the relationship between the NHS and its users asserting that the NHS ‘will place greater emphasis on the outcomes of treatment and care. It will focus on things that really matter' (Department of Health 1997: 66). Potentially dangerous practitioners should be rapidly identified from the evidence of the outcome of their practice and action taken when necessary to minimize harm.

The limits of encoded knowledge

There are some challenges in developing an evidence-based health care system. While encoding knowledge may be designed to structure and control professional decision-making, it is not clear it can achieve this. In particular it appears that encoded knowledge needs to be interpreted and appropriately applied, i.e. it requires other forms of knowledge. While there are few areas of healthcare in which there has been a complete shift to encoded knowledge, in many areas of practice there is increased use of computer-based systems designed to ‘support’ clinical decision-making. Such systems provide a new context and set of resources but they do not remove judgement or negotiation from decision-making. Prior and his colleagues examined the ways in which clinicians used Cyrillic, a computer-based programme, to estimate patients' risk of cancer. Cyrillic made risk ‘visible’ by using inputted data on relatives to draw a family tree of cancer and providing a numerical estimate of personal risk (Prior et al. 2002: 248). Prior and his colleagues found that clinicians had to make ‘sense’ of results and images and this involved craftwork especially in the laboratory. Such craftwork meant that there was always a large chunk of “tacit knowledge” embedded in professional decision making’ (Prior et al. 2002: 256).

Even where there is a major shift to the use of encoded knowledge, studies have found that such knowledge is not adequate or sufficient as a basis for decision making. In an analysis of two projects designed to promote patient safety and reduce risk by reducing drug errors and preventing falls, Proctor found that developments only took place when the technical developments grounded in the use of encoded knowledge were combined with a recognition of the broader social context, especially engaging patients and carers and professionals recognizing their experiential knowledge (Proctor 2002: 57). Similar findings have come from research on NHS Direct, a telephone advice service in which nurses use knowledge encoded within computerized decision support software or algorithms to advice callers about self-care and use of other services. In an observational study Ruston and her colleagues found that following an initial conversation with a caller nurses often made a judgement on the best course of action, they then checked this against that prescribed by the computer algorithm, overriding the algorithm if there was disagreement (Ruston 2006).

The limits of rationalization

The development of a healthcare system based on the application of encoded knowledge is designed not only to improve outcomes and reduce dangerous practice but also to increase public and patient confidence in the NHS creating a ‘high-trust’ organization (Department of Health 2000: para 6.1). As Taylor-Gooby observes, current reforms in healthcare focus on the development of a system in which patient-consumers are provided with the knowledge needed to make informed decisions:

The reforms address some of the problems of information asymmetry identified in healthcare markets … through the widespread publication of league table results, star-ratings and Commission reports (Taylor-Gooby 2006: 98).

Leaving aside the practical problems involved in developing this system, i.e. whether it really works in terms of providing patient choice and controlling professional decision-making, there is serious flaw in the logic. It emphasizes the instrumental rationality of patients and the public and their acceptance and response to information on the overall performance of the system. It neglects the more personal and emotional ways in which individuals judge and respond to their interaction with health professionals. As Calnan and Rowe note, trust can involve using and processing knowledge to reach a judgement but it can also be more intuitive, based on feelings or emotions:

Trust has been characterized as a multi-layered concept primarily consisting of a cognitive element (grounded on rational and instrumental judgments) and an affective dimension (grounded on relationships and affective bonds generated through interaction, empathy and identification with others) (Calnan and Rowe 2006: 3).

The emotional component of healthcare is likely to be quite high. Not only does healthcare involve intimate personal interventions that have a major impact on personal identity and sense of self (Alaszewski et al. 2006) but individuals often have to make ‘fateful’ life and death decisions in stressful and uncertain situations (Barbalet 2002). As Taylor-Gooby notes, the current health reforms in the UK tend to neglect the personal and emotional aspects of care, especially the extent to which patients feel that they and their values are respected:

The current direction of reform leaves affect and emotion out of consideration. There is no sentiment in business, or in a business plan, however good a business plan it happens to be. The appeal is entirely to reason, both in the way the reforms work (through shifting incentives to change the way those involved behave) and in the understanding of people's perceptions of and responses to public sector institutions which is implicit in them (Taylor-Gooby 2006: 101).

The effect of concentrating on cognitive rationality is leading to a system that functions efficiently in curing patients but which fails to meet the personal and emotional needs of patients and therefore does not inspire confidence and trust.

Conclusion

This editorial traces the move from traditional society, where empirical and other sources of knowledge including social values and beliefs are integrated into a single explanatory system, towards a modern society where an enhanced explanatory potential of natural science is accompanied by increased lay-awareness of the limits of this knowledge and corresponding uncertainty. The agency relationship represents one potential means of reducing the complexity inherent in this uncertainty, though this relationship is highly dependent on trust in the professional. The fallibility of the expert, epitomized in various medical scandals, has lead to an increasingly rationalized healthcare system in the UK where the systemization and monitoring of medical practice is applied in an attempt to assure the quality of medical care, and thus furnish trust in the NHS, through enhanced auditing and accountability of clinical decision making.

This shift towards a ‘machine bureaucracy’ (Flynn 2002) misconstrues the nature of medicine. Schön (1983) emphasizes the artistry required in the successful practice of medicine as opposed to a more technical/rational model. In particular he highlights the recurrent presence of abnormal situations where usual protocol is not sufficient and creative reflection is necessary. Not only does the move to encoded knowledge fail to consider that it is these atypical scenarios, where protocol is inadequate, which are also potentially the riskiest; but in their more everyday application, ‘routines tend to become increasingly dysfunctional over time: not only do they fail to adjust to new circumstances but “shortcuts” gradually intrude, some of which only help professionals to cope with pressure at the expense of helping their clients’ (Eraut 1994: 112). The rationalization of healthcare is thus limited in that good practice is reduced to ‘a matter of convention, efficiency and “internal coherence”’ (Marcuse 1989: 121) and the importance of the patient is lost under the ‘meta-regulation’ (Power 2003: 198) of the system.

Whereas the Azande can combine empirical and metaphysical understandings due to the latter's universal acceptance, the instrumental rationality of clinical governance typifies the way modern science, in having to siphon off norms and emotions due to their lack of generalizable validity (Habermas 1974: 264, 1987), is blind to the human, emotional and communicative dimensions of healthcare which are crucial to positive patient experience and trust.

Risk is being applied in this context as a value-neutral concept, its acceptance achieved by the politicization of fear (Furedi 2005: 132), yet its effect, paradoxically, is the depoliticization of everything else (Pieterman 2005). The ‘heuristics of fear’ (Jonas 1984) have emerged to fill the ideological vacuum left after the demise of left-right political discourse (Giddens 1994) and political attention has thus shifted towards security and risk. Risk-aversion, as an apparent universalized moral basis for policy (Furedi 2005: 137), obstructs political debate in that proposals based on this premise are seemingly indisputable. Yet the failure of policies to recognize the affective dimensions of uncertainty is likely to create a ‘trust deficit,’ a system that may work better but is trusted less.